Last month I cut six podcast trailers. Five of them were fine. One of them was good, and the difference had nothing to do with the edit. It had to do with the fact that the host stopped trying to describe the show and just let a 40 second argument play out with no setup.
That is the whole thesis of this post, so I will say it plainly and then spend the rest of the time defending it.
A trailer that describes your podcast will lose to a trailer that simply is your podcast for one minute.
Most trailers I get sent are brochures read aloud. "Welcome to the show where we explore the intersection of business and creativity, with new guests every week." Nobody has ever subscribed to anything because of that sentence. It contains zero information. Swap three nouns and it describes 40,000 other shows.
Why most podcast trailers fail before the edit starts
The failure is upstream of editing. Hosts record a trailer the way they would record an about page: mission, promise, cadence, sign off. Then they hand it to us and hope the music fixes it.
Music does not fix it. I have tried. You can put a great bed under a flat script and all you get is a flat script with a soundtrack. When a host asks me why the trailer is not converting, my first question is never about the mix. It is: what actually happens in this trailer? Usually the answer is nothing happens. Someone talks about things that might happen later.
Here is the second failure, and it is quieter. Trailers get made once, at launch, and then never touched again. A show at episode 90 is a completely different show than it was at episode 1, but the trailer is still promising the thing the host thought they were making two years ago. If you are already rethinking your positioning to the point where you are considering a new name for the show, your trailer is definitely stale too. The name and the trailer are the same decision wearing different clothes.
The 45 second rule, and why I keep breaking it upward
The internet consensus is that a podcast trailer should be 60 to 90 seconds. I think that is roughly right and slightly cowardly.
What I actually watch in the data is drop off, and drop off almost never happens because a trailer was long. It happens because the trailer was boring at second nine. A 2 minute trailer that opens with a genuinely strange sentence will outperform a 45 second trailer that opens with a mission statement, every time.
So my working rule now:
- 0 to 8 seconds: something is already happening. A real line from a real episode, mid thought. No music swell first, no "hey everyone".
- 8 to 25 seconds: the tension. Why this conversation is not settled. Ideally two voices disagreeing, or one voice admitting something.
- 25 to 50 seconds: the range. Three or four fast cuts that prove the show goes to more than one place.
- 50 to 70 seconds: the only part where the host says what the show is and how often it comes out.
- Last 5 seconds: name of show, and one instruction. One.
Notice the description lives at the end, not the beginning. That inversion is the single highest leverage change I make to trailers, and it costs nothing.
If your trailer could still make sense with the audio muted and the text read as a paragraph, you did not make a trailer. You made a press release with a fade in.
Pull from real episodes, always
I will fight anyone on this. The clips in your trailer should come from published episodes, not from a special trailer recording session.
Two reasons. First, a purpose recorded trailer is performed, and people hear performance instantly. The energy is one notch too high and the room tone does not match. Second, and more practically, real episode clips are honest advertising. If someone subscribes because of a clip that represents the actual show, they stay. If they subscribe because of a slick promo that oversells, they churn on episode one and you have bought yourself a worse retention number.
This is where format matters more than people admit. A narrative show and a conversation show need structurally different trailers, and the difference is not cosmetic. I wrote more about that in the piece on narrative and conversation formats, but the short version is: narrative trailers can tease a story question, conversation trailers cannot. Conversation shows have to sell chemistry, which means you need two voices overlapping, interrupting, laughing. One voice alone never sells a two person show.
The clip selection pass nobody wants to do
Here is the unglamorous part. Finding trailer clips means someone has to actually listen to the episodes.
Not scan the transcript. Listen. The best trailer moments are almost never the best written moments, because what makes them work is delivery, a pause, a laugh landing half a beat late. A transcript search for interesting keywords will hand you a paragraph that reads well and plays dead.
This is why our editors watch the whole episode instead of letting software auto cut by silence detection. Silence detection has no opinion about which pause is dead air and which pause is the entire joke. On a trailer, that distinction is the job.
When we build a trailer from a back catalogue, the pass looks like this:
- Pull 20 to 30 candidate moments across the last 8 to 10 episodes.
- Cut them down to 15 to 25 seconds each, cold, no context.
- Play them to someone who has never heard the show. Watch their face.
- Keep the four that got a reaction. Throw out the ones you personally love that got nothing.
Step four is where hosts lose their nerve. Your favourite moment is usually favourite because you remember the conversation around it. A stranger does not have that. Trailers are made entirely for strangers.
Where the trailer actually gets seen
A trailer is not one asset. This is the part most shows get wrong operationally.
The audio trailer sits in your feed as a pinned episode and does a modest job. The version that actually recruits listeners in 2026 is vertical, 30 to 50 seconds, captioned, and living on the platforms where people scroll. Same edit philosophy, tighter, and the show name has to appear as text in the first two seconds because most people never hear the audio at all.
If you are already producing vertical clips from episodes, the trailer is just your best performing clip logic applied to your best four moments instead of one. We charge $25 per vertical clip as an add on, and for a trailer relaunch I usually suggest three cuts: one hook forward, one that leads with the disagreement, one that leads with the funniest line. Then let the platform tell you which one is right, because your instinct and the platform will disagree and the platform will be correct.
For YouTube specifically, the trailer and the thumbnail are a single unit. A perfect 60 second trailer under a lazy thumbnail gets 200 views. Our sister studio ClickTheory does thumbnails at 8 for $250, and I mention it here because I have watched the exact same trailer double its click through rate with nothing changed except the frame in front of it.
When to make a new trailer, and when not to
Not every show needs a trailer refresh. Here is my honest filter.
Make a new trailer if: you have changed format, changed co host, changed the core subject, come back from a break longer than two months, or crossed roughly episode 50 with a trailer made before episode 5.
Do not bother if: you are five episodes in and still figuring out the show. Your trailer will be wrong by episode 20 and you will have spent effort on a moving target. Publish more episodes first.
Shows that run in seasons rather than publishing continuously get an advantage here, because a season is a natural trailer moment. New season, new trailer, and the trailer can be genuinely specific about what this run of episodes is about instead of making vague forever promises. Continuous shows have to manufacture that moment deliberately, usually once a year.
The production side, briefly
I will keep this short because it is not the interesting part.
A trailer built from existing episodes is a small job. If you are already sending us episodes on Starter or Studio, a trailer is typically an add on we schedule alongside a normal episode rather than a separate project. Our regular turnaround is a first cut in 48 hours with revisions the next working day, and trailers usually take one more round than an episode does, because trailers are opinionated and opinions need arguing about.
Recording quality matters more on a trailer than anywhere else, because the trailer is the first audio a stranger hears from you. If your remote recordings are inconsistent, fix that before you make the trailer. The comparison I keep pointing people to is Riverside against Zoom for podcast recording, and the summary is that a trailer cut from local recordings sounds like a show worth subscribing to, while a trailer cut from compressed call audio sounds like a webinar.
One more thing. Do not put four minutes of custom music score under a trailer for a show that has no music anywhere else. It sets an expectation the episodes will not meet. I have turned down trailer briefs for this reason, which is a longer conversation I got into in the post about why I turn down some podcast clients.
What I would do if I were starting from zero today
Record ten episodes. Do not make a trailer. Look at which episode people finished. Pull the four moments that made you sit up when you listened back. Cut them together with almost no glue, add eight seconds of you saying what the show is and when it comes out, and put the show name on screen for the vertical version.
That is it. No script. No voiceover artist. No mission statement.
The trailer is not where you explain yourself. It is where you prove you are worth 30 more minutes. Everything else is decoration, and decoration has never once made me subscribe to a podcast.
If you want a second opinion on whether your current trailer is doing any work, send it over. I will tell you straight, and if the answer is that it is fine, I will say that too.